LLMsBook is a practical code summary repository for the book "Large Language Model Practical Guide: Application Practice and Scenario Implementation," hosted on GitHub. The project organizes code by book chapters and provides practical examples of multiple large language model applications and fine-tuning scenarios.
Purpose: This project is mainly used to support learning of large language model application practice and scenario implementation alongside the book, helping readers understand related technologies through runnable code examples.
Main capabilities (by chapter):
- Chapter 2: tokenizer training methods and PEFT-based Llama model fine-tuning practice.
- Chapter 3: DPO-based preference alignment practice.
- Chapter 5: DeepSeek Coder-based SQL fine-tuning practice.
- Chapter 6: Baichuan-based role-playing model fine-tuning practice.
- Chapter 7: Qwen-based dialogue element extraction model fine-tuning practice.
- Chapter 9: BGE model-based vector fine-tuning practice; ChatGLM3-based knowledge base answer generation task fine-tuning practice; Streamlit-based knowledge base answer application building.
- Chapter 10: Langchain-based AutoGPT building practice.
Usage: The repository is a code summary; users can clone or download the code and study and run it according to the corresponding chapter content. Specific installation commands, dependency configuration, and running steps are not currently provided in the supplied materials.
Applicable scenarios: Suitable for developers and learners who are reading the book or wish to learn practical technologies such as large language model fine-tuning, preference alignment, knowledge base question answering, and AutoGPT building.
Limitations: The materials do not provide specific environment requirements, dependency versions, running instructions, license information, or pricing details; this content is not currently provided. The repository also records book errata, mainly correcting typos, clerical errors, and ambiguous parts, and welcomes feedback via issues. The contact email is logcongcong@gmail.com.
